Final_Report

h1b_df <- DATA

SECTION 1: BACKGROUND

SLIDE 7: MAP

htmltools::includeHTML("rMaps.html")

SECTION 2: STEM

SLIDE 9: Stem and non-stem histogram

DATA %>%
  group_by(year, stem) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n, fill = as.factor(stem))) +
  geom_bar(stat = "identity", position = position_stack(reverse = FALSE), alpha = 0.8) +
  scale_fill_manual(values = c("#00B2E4", "#0F2A48"), name = "", labels = c("non-STEM", "STEM")) +
  labs(y="\n Number", x="\n Year") + 
  scale_y_continuous(labels = function(x){paste0(x/1000, 'K')}) + 
  theme_minimal() + 
  theme(plot.title = element_text(hjust=0.5)) +
  theme(legend.position = "top")

SLIDE 10: STEM and non-STEM boxplot: “Prevailing wage boxplot throughout years”

DATA %>%
  filter(prevailing_wage >= 0 & prevailing_wage <= 600000) %>%
  ggplot(aes(x = as.factor(year), y = prevailing_wage)) +
  geom_boxplot(aes(fill=as.factor(stem)), outlier.shape = NA, alpha=0.9, color = "white") +
  theme_minimal() +
 labs(x = "\n Prevailing wage ($)", y = "Density \n") +
  scale_y_continuous(labels = function(x){paste0(x / 1000, "K")}) +
  scale_color_manual(values = c("#00B2E4", "#0F2A48"), name = "", labels = c("non-STEM", "STEM")) +
  scale_fill_manual(values = c("#00B2E4", "#0F2A48"), name = "", labels = c("non-STEM", "STEM")) +
  coord_cartesian(ylim = c(40000, 100000))

SLIDE 11: Stem and non-stem prevailing wage distribution in 2018

DATA %>%
  filter(year == 2018) %>%
  filter(prevailing_wage >= 0 & prevailing_wage <= 400000) %>% 
  ggplot(aes(x = prevailing_wage, color = as.factor(stem), fill = as.factor(stem))) +
  geom_density(adjust = 2, alpha = 0.7)+
  theme_classic() +
  labs(x = "\n Prevailing wage ($)", y = expression(Density~(10^{-3})), title = "Distribution of prevailing wage in 2018") +
  scale_x_continuous(labels = function(x){paste0(x/1000, 'K')}) +
  scale_y_continuous(labels = function(x){paste0(x * 1000, "")}) +
  scale_color_manual(values = c("#00B2E4", "#0F2A48"), name = "", labels = c("non-STEM", "STEM")) +
  scale_fill_manual(values = c("#00B2E4", "#0F2A48"), name = "", labels = c("non-STEM", "STEM")) +
  theme(plot.title = element_text(hjust=0.5)) +
  theme_minimal()

SLIDE 12 AND 13: bar plot: stem jobs

dt <- DATA %>% 
  mutate(job_title = ifelse(str_detect(job_title, "business")& (str_detect(job_title, "analyst")|str_detect(job_title, "intelligence")) , "business analyst", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& (str_detect(job_title, "engineer")|str_detect(job_title, "warehouse")), "data engineer", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& str_detect(job_title, "scientist"), "data scientist", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& str_detect(job_title, "analyst|analytics"), "data analyst", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "machine|deep")& str_detect(job_title, "learning"), "deep learning & machine learning", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "product") & str_detect(job_title, "analyst|engineer|data"), "data product analyst", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "senior systems analyst jc60"), "senior systems analyst", job_title)) %>% 
  filter(stem == 1)
# 
# dt %>% 
#   filter(year == 2017) %>% 
#   group_by(job_title) %>% 
#   summarise(n = n()) %>%
#   arrange(-n) %>% 
#   slice(1:20) %>% 
#   mutate(cat = ifelse(str_detect(job_title, "analyst|data"), "data", "non-data")) %>%
#   arrange(n) %>% 
#   ggbarplot("job_title", "n", fill = "cat", palette = c("#FC2967", "#00B2E4"), alpha = 0.8, orientation = "horiz", color = "white") +
#   labs(y = "Number of Aplications", x = "Job Titles") +
#   theme_minimal()

dt %>% 
  filter(year == 2018) %>% 
  group_by(job_title) %>% 
  summarise(n = n()) %>%
  arrange(-n) %>% 
  slice(1:20) %>% 
  mutate(cat = ifelse(str_detect(job_title, "analyst|data"), "data", "non-data")) %>%
  arrange(n) %>% 
  ggbarplot("job_title", "n", fill = "cat", palette = c("#FC2967", "#00B2E4"), alpha = 0.8, orientation = "horiz", color = "white") +
  labs(y = "Number of Aplications", x = "Job Titles")+
  theme_minimal()+
  theme(
    legend.position = "none"
  ) +
  theme(panel.grid.major.y = element_blank())

# original
dt %>% 
  filter(year == 2018) %>% 
  group_by(job_title) %>% 
  summarise(n = n()) %>%
  arrange(-n) %>% 
  slice(1:20) %>% 
  mutate(cat = ifelse(str_detect(job_title, "analyst|data"), "data", "non-data")) %>%
  arrange(n) %>% 
  ggbarplot("job_title", "n", fill = "#00B2E4", alpha = 0.8, orientation = "horiz", color = "white") +
  labs(y = "Number of Aplications", x = "Job Titles") +
  theme_minimal()+
  theme(panel.grid.major.y = element_blank())

Slide 15

dt <- DATA %>% 
  filter(data_job_title %in% c("Business Analyst","Data Engineer","Data Analyst", "Data Scientist")) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  arrange(-n) 

p1 <- dt %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2)+
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "Applications Data related jobs", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") + 
  theme_minimal() +
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))
p1

slide 16

filtered_job <- DATA %>% 
  filter(data_job_title %in% c("Business Analyst","Data Engineer", "Data Scientist","Data Analyst"))


filtered_job$data_job_title <- factor(filtered_job$data_job_title, levels = c("Data Analyst", "Business Analyst", "Data Engineer", "Data Scientist"), ordered = TRUE)


filtered_job %>% group_by(data_job_title) %>% summarise(mean = mean(prevailing_wage))

filtered_job %>%  
  ggplot(aes(y = prevailing_wage, x = data_job_title)) +
  geom_boxplot(fill = "#00B2E4", outlier.shape = NA, color = "white", width = 0.5) +
  coord_cartesian(ylim = c(50000, 120000)) +
  theme_minimal() +
  labs(x = "\n Job title", y = "Prevailing wage (per year) \n", title = "Prevailing wages in data related jobs") +
  theme(plot.title = element_text(size = 1),
        text = element_text(size = 12),
        axis.title = element_text(size = 16),
        axis.text.x= element_text(size = 10),
        axis.text.y = element_text(size = 10)) +
  theme(plot.title = element_text(hjust = 0.5))

Slide 17

top_com <- DATA %>% 
  filter(data_job_title %in% c("Business Analyst", "Data Analyst", "Data Engineer", "Data Scientist")) %>%
  group_by(employer_name) %>% 
  summarise(n = n()) %>% 
  arrange(-n) %>% 
  slice(1:20) %>% 
  ungroup()

top_com <-top_com$employer_name 

spa_data <- DATA %>% 
  filter(employer_name %in% top_com) %>% 
  filter(employer_name != "capgemini america inc") %>% 
  filter(data_job_title %in% c("Business Analyst", "Data Analyst", "Data Engineer", "Data Scientist")) %>%
  group_by(employer_name, year) %>% 
  summarise(n = n()) 

con=c("infosys limited", "deloitte consulting llp","amazon","facebook", "ibm")
myColor <- c("#465881", "#FFB6C1", "#00909e", "#00B2E4", "#FC2967")

# c(amazon, "deloitte", "#facebook", "ibm", "info")

ggplot(spa_data, aes(x = year, y = n, group = employer_name)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(x = "Year",
       y = "Number",
       col = "") +
  theme_classic()+
  geom_line(data = subset(spa_data, employer_name %in% con),aes(col=employer_name),size=1.2)+
  theme(plot.title = element_text(size = 14),
              text = element_text(size = 12),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11)) +
  scale_color_manual(values = myColor, breaks = c("infosys limited", "deloitte consulting llp","amazon","facebook", "ibm"), labels = c("Infosys Limited", "Deloitte Consulting","Amazon","Facebook", "IBM"))  +
  theme_minimal() +
  theme(legend.position = "right")

Slide 17

Top Tech Companies and Their Trends

soc_top_tech <- c("Apple",
                  "Microsoft",
                  "Amazon",
                  "Facebook",
                  "Google",
                  "IBM")


DATA %>% 
  filter(data_job_title %in% c("Business Analyst","Data Engineer","Data Analyst", "Data Scientist")) %>% 
  mutate(employer_name = ifelse(str_detect(employer_name,'apple'), 'Apple',
                          ifelse(str_detect(employer_name,'microsoft'), 'Microsoft',
                          ifelse(str_detect(employer_name,'amazon'),'Amazon',
                          ifelse(str_detect(employer_name,'facebook'),'Facebook',
                          ifelse(str_detect(employer_name,'google'),'Google',
                          ifelse(str_detect(employer_name, 'ibm'),'IBM','no'))))))) %>% 
  group_by(employer_name, year) %>% 
  summarise(tot = n()) %>% 
  filter(!employer_name %in% c("no", NA)) %>% 
  ungroup()%>% 
  ggplot()+
  scale_color_brewer(palette = "RdBu") +
  geom_line(aes(x = year, y = tot, color = employer_name), size = 1.5)+
  labs( y = "Number of Applications", x = "Year", color = "Employer")+
  theme_minimal()

Data jobs in top companies

filtered_job_all <- DATA %>% 
  filter(data_job_title %in% c("Business Analyst","Data Analyst", "Data Engineer", "Data Scientist")) %>% 
  mutate(employer_name = ifelse(str_detect(employer_name,'apple'), 'Apple',
                          ifelse(str_detect(employer_name,'microsoft'), 'Microsoft',
                          ifelse(str_detect(employer_name,'amazon'),'Amazon',
                          ifelse(str_detect(employer_name,'facebook'),'Facebook',
                          ifelse(str_detect(employer_name,'google'),'Google',
                          ifelse(str_detect(employer_name, 'ibm'),'IBM',employer_name))))))) 
  

#IBM

top_com <- DATA %>% 
  filter(data_job_title %in% c("Business Analyst", "Data Analyst", "Data Engineer", "Data Scientist")) %>%
  group_by(employer_name) %>% 
  summarise(n = n()) %>% 
  arrange(-n) %>% 
  slice(1:20) %>% 
  ungroup()

top_com <-top_com$employer_name 

ibm_data <- DATA %>% 
  filter( str_detect(employer_name, "ibm")) %>% 
  group_by(job_title, year) %>% 
  summarise(n = n())%>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title,"business")&(str_detect(job_title,"analyst")|str_detect(job_title, "intelligence")), 
                  "Business Analyst", 
                  NA)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "data")&(str_detect(job_title,"engineer")|str_detect(job_title, "warehouse")),
                  "Data Engineer", 
                  data_job_title)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "data")&str_detect(job_title, "scientist"), 
                  "Data Scientist", 
                  data_job_title)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "data")&str_detect(job_title, "analyst|analytics"), 
                  "Data Analyst", 
                  data_job_title)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "machine|deep")&str_detect(job_title, "learning"), 
                  "Data Scientist", 
                  data_job_title)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "product")&str_detect(job_title, "analyst|engineer|data"), 
                  "Data Analyst", 
                  data_job_title)) %>% 
  mutate(data_job_title = 
           ifelse(str_detect(job_title, "product")&str_detect(job_title, "manager"), 
                  "Product Manager", 
                  data_job_title))

con=c("infosys limited", "deloitte consulting llp","amazon","facebook", "ibm")
myColor <- c("#465881", "#FFB6C1", "#00909e", "#00B2E4", "#FC2967")

# c(amazon, "deloitte", "#facebook", "ibm", "info")

ggplot(ibm_data, aes(x = year, y = n, group = job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(x = "Year",
       y = "Number",
       col = "") +
  theme_classic()+
  geom_line(data = subset(ibm_data, data_job_title %in% c("Business Analyst", "Data Analyst", "Data Engineer", "Data Scientist")),aes(col=data_job_title),size=1.2)+
  theme(plot.title = element_text(size = 14),
              text = element_text(size = 12),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11),
        legend.position = "top",)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  ylim(0,150)

filtered_job_all %>% 
  filter(employer_name == 'IBM') %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

NA
NA

Infosys

filtered_job_all %>% 
  filter(str_detect(employer_name, 'infosys')) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        #axis.text = element_text(size = 15),
        #axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Accenture

filtered_job_all %>% 
  filter(str_detect(employer_name, 'accenture')) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Microsoft

filtered_job_all %>% 
  filter(employer_name == 'Microsoft') %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Amazon

filtered_job_all %>% 
  filter(employer_name == 'Amazon') %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Deloitte

filtered_job_all %>% 
  filter(str_detect(employer_name, "deloitte")) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Facebook

filtered_job_all %>% 
  filter(str_detect(employer_name, "Facebook")) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Apple

filtered_job_all %>% 
  filter(str_detect(employer_name, "Apple")) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Google

filtered_job_all %>% 
  filter(str_detect(employer_name, "Google")) %>% 
  group_by(data_job_title, year) %>% 
  summarise(n = n()) %>% 
  ggplot(aes(x = year, y = n))+
  #geom_point(aes(color = data_job_title), size = 2) +
  geom_line(aes(color = data_job_title),size = 1.2)+
  labs( y = "", x = "Year", title = "", face = "bold", size = 14)+
  scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme_minimal()+
  theme(legend.position = "top",
        axis.text = element_text(size = 15),
        axis.title.x = element_text(size = 15)
        ) +
  theme(plot.title = element_text(hjust=0.5))

Facebook

h1b_facebook <- h1b_df %>% 
  filter(employer_name == 'facebook') %>% 
  group_by(data_job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)
  

con=c("Data Scientist","software engineer","Data Engineer",'research scientist')
ggplot(h1b_facebook, aes(x = year, y = n, group = data_job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "H1B in Facebook",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_light()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_facebook, data_job_title %in% con),aes(col=data_job_title),size=0.7)+
  theme(plot.title = element_text(size = 14, family = "Helvetica", face = "bold"),
              text = element_text(size = 12, family = "Helvetica"),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11))+
  scale_color_discrete(labels = c("Data Engineer", "Data Scientist", "Research Scientist","Software Engineer"))

Amazon

h1b_amazon <- h1b_df %>% 
  filter(str_detect(employer_name,'amazon')) %>% 
  ungroup(data_job_title) %>% 
  mutate(job_title=ifelse(str_detect(data_job_title,'software development engineer'), 'software development engineer', 
                   ifelse(str_detect(data_job_title,'technical program manager'), 'technical program manager',data_job_title)))%>% 
  group_by(data_job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)

con=c("Data Scientist","software development engineer","business analyst","Data Engineer","technical program manager")
ggplot(h1b_amazon, aes(x = year, y = n, group = data_job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "H1B in Amazon",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_light()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_amazon, data_job_title %in% con),aes(col=data_job_title),size=0.7)+
  theme(plot.title = element_text(size = 14, family = "Helvetica", face = "bold"),
              text = element_text(size = 12, family = "Helvetica"),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11)) +
  scale_color_discrete(labels = c("Business Analyst", "Data Engineer", "Data Scientist","Software Development Engineer","Technical Program Manager"))

Microsoft

h1b_micro <- h1b_df %>% 
  filter(str_detect(employer_name,'microsoft')) %>% 
  ungroup(data_job_title) %>% 
  mutate(job_title=ifelse(str_detect(data_job_title,'software engineer|software development engineer'), 'software engineer', 
                   ifelse(str_detect(data_job_title,'program manager'), 'program manager',data_job_title))) %>% 
  group_by(data_job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)
  

con=c("Data Scientist","software engineer","program manager")
ggplot(h1b_micro, aes(x = year, y = n, group = data_job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "H1B in Microsoft",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_light()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_micro, data_job_title %in% con),aes(col=data_job_title),size=0.7)+
  theme(plot.title = element_text(size = 14, family = "Helvetica", face = "bold"),
              text = element_text(size = 12, family = "Helvetica"),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11)) +
  scale_color_discrete(labels = c("Data Scientist","Program Manager","Software Engineer"))

Google

h1b_google <- h1b_df %>% 
  filter(str_detect(employer_name,'google')) 
  

h1b_google <- h1b_google %>% 
  ungroup(job_title) %>% 
  mutate(job_title=ifelse(str_detect(job_title,'software engineer'), 'software engineer',
              ifelse(str_detect(job_title,'hardware engineer'), 'hardware engineer',job_title))) %>% 
  group_by(job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)

con=c("Data Scientist","software engineer","program manager","product manager","business analyst")
ggplot(h1b_google, aes(x = year, y = n, group = job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "H1B in Google",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_light()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_google, job_title %in% con),aes(col=job_title),size=0.7)+
  theme(plot.title = element_text(size = 14, family = "Helvetica", face = "bold"),
              text = element_text(size = 12, family = "Helvetica"),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11)) +
  scale_color_discrete(labels = c("Business Analyst", "Data Scientist","Product Manager","Program Manager", "Software Engineer"))

IBM

h1b_ibm <- h1b_df %>% 
  filter(str_detect(employer_name,'ibm')) 
  

h1b_ibm <- h1b_ibm %>% 
  group_by(data_job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)

con=c("Business Analyst","Data Engineer","Data Scientist","Data Analyst")
ggplot(h1b_ibm, aes(x = year, y = n, group = data_job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_minimal()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_ibm, data_job_title %in% con),aes(col=data_job_title),size=1.1)+
    scale_color_manual(values=c("#FC2967","#465881", "#00909e", "#00B2E4"), name = "") +
  theme(plot.title = element_text(size = 14, family = "Helvetica"),
              text = element_text(size = 12, family = "Helvetica"),
              #axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11),
        legend.position = "none")+
  coord_cartesian(ylim=c(0,1000))

Uber

h1b_uber <- h1b_df %>% 
  filter(str_detect(employer_name,'uber')) %>% 
  ungroup(data_job_title) %>% 
  mutate(data_job_title=ifelse(str_detect(data_job_title,'software engineer'), 'Software Engineer', 
        data_job_title)) %>% 
  group_by(data_job_title,year) %>% 
  summarise(n=n()) %>% 
  arrange(-n)

con=c("Software Engineer","Data Scientist","Data Analyst")
ggplot(h1b_uber, aes(x = year, y = n, group = data_job_title)) +
  geom_line(alpha = 0.3, 
            size = 0.2) +
  labs(title = "H1B in Uber",
       x = "Year",
       y = "Number",
       col = 'Job Title') +
  theme_light()+
  theme(plot.title = element_text(hjust = 0.5))+
   geom_line(data = subset(h1b_uber, data_job_title %in% con),aes(col=data_job_title),size=0.7)+
  theme(plot.title = element_text(size = 14, family = "Helvetica", face = "bold"),
              text = element_text(size = 12, family = "Helvetica"),
              axis.title = element_text(face="bold"),
              axis.text.x=element_text(size = 11))

Alluvia

filtered_job <- DATA %>% 
  filter(year == 2018) %>% 
  filter(data_job_title %in% c("Business Analyst","Data Analyst","Data Engineer", "Data Scientist"))


top_10_state <- filtered_job %>% 
  group_by(worksite_state_abb) %>% 
  summarise(count = n()) %>% 
  arrange(desc(count)) %>% 
  slice(1:10) 

flows <- filtered_job %>% 
  filter(worksite_state_abb %in% top_10_state$worksite_state_abb) %>% 
  group_by(data_job_title,
           worksite_state_abb) %>%
  summarise(count = n())


ggplot(flows, aes(y =count, axis1=data_job_title, axis2=worksite_state_abb)) +
  geom_alluvium(aes(fill = data_job_title)) +
  geom_stratum(width = 1/8, fill = "#0F2A48", color = "grey", alpha = 0.9) +
  geom_label(stat = "stratum", label.strata = TRUE) +
  labs(x="",y="")+
  theme_minimal()+
  theme(legend.position = "top") +
  scale_fill_manual(values =c("#FC2967","#465881", "#00909e", "#00B2E4"), name=NULL)+
  scale_x_discrete(limits = c("Job title", "State"), expand = c(.05, .05)) +
  theme(axis.text.x=element_text(size = 12, face = "bold"))+
  theme(line = element_blank())

bar plot:stem program in different job titles

DATA %>% 
  mutate(job_title = ifelse(str_detect(job_title, "business")& (str_detect(job_title, "analyst")|str_detect(job_title, "intelligence")) , "business analyst", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& (str_detect(job_title, "engineer")|str_detect(job_title, "warehouse")), "data engineer", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& str_detect(job_title, "scientist"), "data scientist", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "data")& str_detect(job_title, "analyst|analytics"), "data analyst", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "machine|deep")& str_detect(job_title, "learning"), "deep learning & machine learning", job_title)) %>% 
  mutate(job_title = ifelse(str_detect(job_title, "product") & str_detect(job_title, "analyst|engineer|data"), "data product analyst", job_title)) %>% 
  filter(year == 2018) %>% 
  filter(stem == 1) %>% 
  group_by(job_title) %>% 
  summarise(n = n()) %>%
  arrange(-n) %>% 
  slice(1:20) %>% 
  ggbarplot("job_title", "n", fill = "#00B2E4", color = "white", alpha = 0.8, sort.val = "asc", orientation = "horiz") %>% 
  labs(title = "stem program")

[[1]]

$title
[1] "stem program"

attr(,"class")
[1] "labels"

MAP - data job with the highest number in each city

ggplot(points) + 
  geom_sf(data = state_maps, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon,100), y = jitter(lat,100), color = max_job, size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        legend.box = "horizontal",
        legend.position = c(0,-0.1), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")

ggplot(points_ca) + 
  geom_sf(data = state_maps_ca, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon), y = jitter(lat), color = max_job, size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        #legend.position = c(0,0), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")

ggplot(points_tx) + 
  geom_sf(data = state_maps_tx, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon), y = jitter(lat), color = max_job, size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        #legend.position = c(0,0), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")

ggplot(points_northeast) + 
  geom_sf(data = state_maps_northeast, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon), y = jitter(lat), color = max_job, size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        #legend.position = c(0,0), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")

ggplot(points_midwest) + 
  geom_sf(data = state_maps_midwest, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon,100), y = jitter(lat,100), color = max_job, size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        #legend.position = c(0,0), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")

ggplot(points_tn) + 
  geom_sf(data = state_maps_tn, color = "white", fill = "lightgrey",alpha=0.5) +
  geom_point(aes(x = jitter(lon), y = jitter(lat), color = max_job,size = max), alpha = 0.5) + 
  scale_color_manual(values = c("Business Analyst" = "#00B2E4", 
                                "Data Analyst" = "#0F2A48",
                                "Data Engineer" = "#008B8B",
                                "Data Scientist" = "#FC2967"),
                     name = "Job Title") +
  theme(panel.background = element_blank(),
        legend.background = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        #legend.position = c(0,0), legend.justification = c(0,0)
        )+
  labs(x = "", y = "", title = "", size="Number")